DE102005057406A1 - Method for recording a sound source with time-variable directional characteristics and for playback and system for carrying out the method - Google Patents

Abstract

The invention relates to a method for recording sound signals from one or more sound sources arranged in a recording room with temporally variable directional characteristics and for perceptually reproducing the sound signals and the directional information of the sound sources in a playback room; the invention also relates to a system for carrying out the method. In order to be able to record, transmit and reproduce the directional information of a sound source in real time, only the main emission direction of the sound signals emitted by the sound source as a function of time is recorded in the recording room and the reproduction is carried out as a function of the detected main emission direction. To convey the directional information, the sound signals are reproduced with a first reproduction unit assigned to the sound source and at least one second reproduction unit arranged at a distance from the first reproduction unit. The playback with the second playback unit or units takes place with time delays gegenüber compared to the first playback unit.

Description

The The invention relates to a method for recording sound signals of a arranged in a recording room sound source with temporally variable directional characteristics and for reproducing the sound signals in a playback room, as well as a system for performing the Process.

It Various methods are known with which the attempt is made Sound impression that arises in a room to record and reproduce. The best known method is the stereo method and the developments of which the location of a sound source is detected during recording and is displayed during playback. When playing there but only a limited Area in which the location of the recorded sound source is reproduced correctly becomes. Other reproduction methods that synthesize the recorded sound field such as the wave field synthesis, however, the location of the sound source independently from the listener position play correctly.

at none of these methods are time - variable information on the Radiation direction of a sound source recorded and reproduced. Are sound sources with time-varying directional characteristics recorded, information is lost. For example, for the transmission of a Videoconferencing in which a participant approaches different Contact participants and address them specifically, be with the known methods do not capture this directional information, recorded and played.

Of the Invention is based on the object, a method for recording, transmission and reproduce sound, with which the information-bearing Properties of the sound sources are reproduced perceptually can.

The Task is solved by a method for receiving sound signals from one in a recording room arranged sound source with time-varying directional characteristic with sound recording means and for reproducing the sound signals with sound reproduction means in a reproduction room, characterized in that the Main emission of the sound emitted from the sound source dependent on is recorded by the time and playback in dependence the detected main emission takes place.

A Sound source with time-varying directivity can thereby in particular, be a participant of a videoconference that is itself to turn to other participants and speak in different directions can. The emitted sound signals are recorded and simultaneously recorded their main direction of emission.

The Recording of the sound signals can be done in a conventional manner with microphones as well as with one or more microphone arrays. The means for detecting the main emission can be of any kind. Especially can acoustic means are used. This can be several microphones and / or one or more microphone arrays are used, which and / or phase differences of the signal in different directions capture, from which the main emission by means of a suitable Signal processing can be determined. If the position of acoustic means, their directional characteristics, and / or the position the sound source are known this information when determining the main emission direction be considered by the signal processing accordingly. Likewise also knowledge about the geometry of the environment and the related ones Properties for the sound propagation, as well as reflection properties in the Determination of the main radiation direction to be included. Especially can be advantageous also information about the measured, approached or simulated directional characteristic of the sound source in the determination the main emission direction. This is especially true if the main radiation direction is to be determined only roughly, what kind of many applications is sufficient.

to However, detection of the main emission direction can also be optical means be used, such as. B. a video capture with pattern recognition. at the participants of a videoconference can be assumed that the speech direction corresponds to the viewing direction. Thus, through a pattern recognition can be determined, in which direction a participant looks, and from this the speech direction can be determined. Especially can also be a combination of acoustic and optical means be used with appropriate signal processing. Possibly can the acoustic means for the recording of the sound signals at the same time for recording the main emission direction are used, and vice versa.

Often it is sufficient to roughly determine the main emission direction. A gradation in 3 or 5 stages, z. B. straight, right and left or straight, at an angle right, right, oblique left and left can be fully sufficient to get the essential information to convey.

The main emission direction can advantageously be the main emission direction in that frequency range which is information-bearing. For this purpose, the frequency range used to determine the main emission direction can be restricted the, z. By the use of a frequency filter.

The Playback of the sound signals should depend on the detected main direction of radiation respectively. Here, the directional radiation of the original source be simulated. This can either be directed by a genuine one Emitted by the sound signal or by a simulated directional playback, that of the listener is perceived as directed playback without being physically directed in the conventional To be meaning. The applicable procedures differ under in the accuracy with which the directional characteristic is reconstructed becomes. In practice, the perception-oriented naturalness the reconstruction or simulation crucial. Hereinafter all such procedures are summarized under the term "directed reproduction" become.

at the method according to the invention can playback the sound signals associated with one of the sound source first playback unit and at least one spaced from the First playback unit arranged second playback unit done. In this case, the position of the first playback unit in the playback room correspond to a virtual position of the sound source in the playback room. The second and second display units can be used to convey the direction information the sound reproduction serve. Preferably, this will be two second Playback units used, one of which on the one and the other on the other side of the first sound playback unit can be positioned. Instead of a second playback unit can be used on each side of the first sound playback unit, respectively also a plurality of spaced apart second display units are arranged, preferably in each case two second display units.

The in the recording room recorded sound signals of the sound source can in Play room from a first playback unit, such. B. one Speakers to be played. This speaker can do that be placed in the playback room that he is at the virtual Position of the sound source is in the playback room. The sound source So it is brought into the playback room, so to speak. The first Playback unit can also be used with multiple speakers, with a set of speakers or with a speaker array become. For example, it is possible the first playback unit by wave field synthesis as a point source to place at the virtual position of the sound source in the playback room, so that the sound source virtually brought in the playback room becomes. This is advantageous for. For video conferencing, where preferably the impression of an actual Conference in the presence of all participants. Sound source would be then a participant in the recording room. The playback would be over one first playback unit, which at the point in the playback room would be placed at the participant in the recording room virtually in the playback room would be present.

The information about the emission direction can be mediated by the fact that the reproduction with the second display units or with respect to the first reproduction unit with a time delay τ with respect to the first reproduction unit he follows. The time delay can do this for the each second playback units to be different. It has revealed that the human ear information about the Radiation direction of a sound source can be taught thereby that some sort of echo or reflection of the sound signal from one or more emitted a plurality of spaced sound sources with a small time delay becomes. The time delay to participant positions where z. In a video conference a participant can be placed between 2 ms and 100 ms so that the echo or reflection is not as separate sound event is processed. The time delay τ of the second Reproducing unit or units may therefore preferably be chosen that the actual Time Delay between the audio signals at least in subregions of the playback room between 2 ms and 100 ms, preferably between 5 ms and 80 ms and in particular between 10 ms and 40 ms.

The Playback by the second playback unit (s) may be performed accordingly the spatial characteristics of the playback room with a reduced Level, in particular one by 1 to 6 dB and preferably in order 2 to 4 dB reduced level done. According to the simulated Directional characteristics, the sound signal before playback through the second (n) playback unit (s) also with a frequency filter be edited, such as a high-pass, low-pass or Bandpass filter. The parameters of the frequency filters can either be fixed or dependent be controlled by the main emission direction.

The second (s) playback unit (s), as well as the first playback unit, one or more speakers or a virtual source with a group of speakers or a speaker array is generated for example by wave field synthesis, be.

For reproduction of the information about the emission direction of a sound source that is as true to the perception as possible, the reproduction level of the first and second reproduction units can also be dependent on the directional characteristic to be simulated be adapted. The reproduction levels are set for this purpose so that the perceptible loudness differences resulting from the directional characteristic are approximated correspondingly at different listener positions. The reproduction levels of the individual reproduction units determined in this way can be defined and stored for different main emission directions. In the case of a time-varying directional characteristic, the detected main emission direction then controls the reproduction levels of the individual reproduction units.

The Of course, the above-described method can also be applied to several Sound sources are used in the recording room. For playing multiple sound sources With the method described, it is particularly advantageous that Sound signals of the individual to be transmitted Sound sources separated from each other. Therefore are different Method for recording the sound signals conceivable. For the record the sound signals can be assigned to the individual sound sources recording material. The Allocation can be either 1: 1, so that each sound source via a own sound recording means, or so that the sound sources group together into several ones Sound recording means are assigned. The position of a given Time active sound source can be combined with both conventional localization algorithms as well as with video capture and pattern recognition. For simultaneous sound emission from more than one sound source can in a grouping of the sound sources to a sound recording means the Sound signals of the individual sound sources with conventional source separation algorithms such as "Blind Source Separation "," Independent Component Analysis "or" Convolutive Source Separation "from each other be separated. If the position of the sound sources to be recorded is known can be used as a recording device for a group of sound sources also a dynamic directionally selective Microphone array that uses the received audio signals processed according to the given positions and for each sound source summarized separately.

The Detection of the main radiation direction of the individual sound sources can with the same principles as described for a sound source respectively. You can do this the individual sound sources corresponding means are assigned. The assignment can be made such that each sound source is above it own directional means has, or so that the sound sources grouped together into several a directional means assigned. For grouped sound sources the recording is done the main emission direction as with a sound source, if given When only one sound source emits sound. Beams at the same time two or more sound sources sound off, so in the first processing step the direction detecting means first the received signals (For example, sound or video signals) the corresponding sound sources assigned. This can be done with optical means by means of object recognition algorithms respectively. For acoustic means, those with the previously described Sound recording means recorded separately sound signals of the sound sources, for the Assignment of the received signals to the corresponding sound sources be used. In a known position of the sound sources may preferably the transfer function between the sound sources and the acoustic direction sensing Means taken into account as well as the directional characteristics of both the direction-sensing By means of as well as the sound recording medium. Only after the assignment the received signals to the associated sound sources becomes the main emission direction for the determined separate sources of sound, with the same pre-described Procedure for a sound source can be used.

The quality the playback can be improved by the fact that a sound signals Sound source, the recording means not assigned to the sound source or direction sensing means are received by an acoustic echo cancellation or crosstalk suppression (cross talk cancellation) become. The minimization of acoustic reflections and background noise with usual Means can also help to improve the quality of reproduction.

For playback the audio signals, each sound source, a first playback unit be assigned. The assignment can be done either 1: 1, so that every sound source over has its own first playback unit, or so that the sound sources groups each assigned to a plurality of a first playback unit are. Depending on the assignment is the reproduced in the playback room spatial Information more or less precise.

As an alternative to the reproduction technique described above, the reproduction can also take place by wave field synthesis. For this purpose, instead of the commonly used point source, the directional characteristic of the sound source for synthesizing the sound field must be taken into account. The directional characteristic to be used is preferably kept ready in a database for this purpose. The directional characteristic may be, for example, a measurement, an approximation obtained from measurements, or an approximation described with a mathematical function. It is also possible to simulate the directional characteristic by a model, such as by means of directional filters, by means of several elementary sources or a direction-dependent excitation. The synthesis of the sound field with the corresponding directional characteristics is controlled by means of the detected main emission direction, so that the information about the emission direction of the sound source is reproduced as a function of time. Of course, the described method can also be used for several sound sources in the recording room.

Next the two playback techniques described so far can for the directed Playback of the sound signals also known from the prior art Multi-speaker system (multi-speaker display device) used whose playback parameters are also dependent on the time determined main emission direction is controlled. Instead of The control of the playback parameters is also a control a rotatable mechanism conceivable. Are several sources in the Recording space available, in the playback room for each sound source, a multi-speaker system to disposal be put.

For the targeted Playback of the sound signals can also other known from the prior art reproduction method be used, for this purpose their playback parameters according to the dependent the time determined main emission must be controlled.

Of the Another object of the invention is to provide a system with which the record, transmission and perceptual reproduction of the information-carrying properties allows the sound sources becomes.

The Task is solved by a system for receiving sound signals from one or more Sound sources with time-varying directional characteristics with sound recording devices in a recording room and for reproducing the sound signals with Tonwiedergabemitteln in a reproduction room, characterized in that the System Means for detecting the main emission directions of the sound signals emitted as a function of the sound source or sources of time and means for reproducing the transmitted sound signals in dependence the detected directions.

The System can have at least two sound recording units assigned to a sound source for recording the sound signals emitted by this sound source and have their Hauptabstrahlrichtung. Alternatively or in addition to this For example, the system may also include optical means for detecting its main direction of emission exhibit.

medium for detecting the main emission direction z. B. microphones or Microphone arrays or video recording means, in particular with pattern recognition, be.

The Playback of the sound signals can be associated with one of the sound source first playback unit and at least one spaced from the first Playback arranged arranged second playback unit. The position of the first playback unit in the playback room can correspond to a virtual position of the sound source in the playback room.

The Playback with the second or the second playback units can be used for subjective generation of a directed sound radiation compared to the first playback unit with a time delay τ. For several second Playback units can for each an individual time delay chosen become.

The System can z. B. for sound transmission be used in videoconferencing. There are fixed Positions in which participants of the conference are present. In dependence of the subscriber positions, the time delay τ of the second playback unit or units chosen such be that actual Time delay between the sound signals at least at the respective participant positions in the reproduction room between 2 ms and 100 ms, preferably between 5 ms and 80 ms and especially between 10 ms and 40 ms.

The Playback through the first and / or second playback unit (s) can with a reduced level, especially with a 1 to 6 dB, and preferably reduced by 2 to 4 dB level, and / or especially depending from the main emission direction.

It It goes without saying that the system for transmitting the sound signals of a sound source on the transmission the sound signals from several sound sources can be extended. This can be done by simply multiplying the above-described means respectively. Advantageously, however, a reduction of the required Means are made such that certain means several sound sources are assigned on the recording side. Alternatively or additionally, too Playback on the playback side have several assignments. The above for the inventive method described allocation possibilities apply analogously for the system. In particular, the number of sound recording units and / or sound reproduction units of the number of sound sources plus 2 correspond.

Further embodiments of the method and the system are specified in the subclaims.

The invention will be described below with reference to the accompanying drawings described in more detail by selected examples:

1 shows a microphone array;

2A and B illustrate a simplified acoustic method for determining the main emission direction of a sound source;

3 shows the determination of the main emission direction of a sound source by means of a reference sound level;

4 shows a method for direction detection at multiple sound sources in the recording room;

5 shows a method in which a separate direction detecting means is used for each sound source;

6 shows a reproduction method for a sound source having a first reproduction unit and at least a second, spaced reproduction unit;

7A and 7B show various methods for realizing the first and second display units;

8A and 8B show reproduction methods for a sound source having a first reproduction unit and a plurality of second spaced-apart reproduction units;

9 shows a reproduction method for a plurality of sound sources with overlapping first and second reproduction units;

10A and 10B demonstrate a simplified reproduction method for direction detection 5 ,

This in 1 shown microphone array MA is used to detect the main emission of a sound source T in the recording room.

The main emission direction of a sound source T is determined by means of a microphone array MA, that is to say a plurality of individual microphones M which are connected to one another. For this purpose, the sound source T is surrounded with these microphones MA in any desired form, for example ring-shaped, as in FIG 1 shown.

In In a first step, the position of the sound source T with respect to Microphone M is determined so that all distances r between sound source T and microphones M are known. The position of the sound source T can for example, by measurement or with a conventional localization algorithm be determined. It can be for the determination of the position to be advantageous, only the frequency ranges with appropriate filters to take into account that no pronounced preferred direction in terms of sound radiation. This applies in many cases to low frequency ranges, for example, below 500 Hz for speech.

The Main emission of the sound source T can from the detected sound levels be determined on the microphones M, in which case the different sound attenuation as well as transit time differences due to the different distances r between the individual microphones M and the sound source T considered become. For directionally selective microphones M, the directional characteristic can also be used the microphones M in the determination of the main emission direction.

more More directions can be detected by microphones, the more precise it can be the main emission direction can be determined. Conversely, the number the necessary microphones are reduced, (a) when the main emission direction should only be roughly recorded, for example, a gradation in only 3 or 5 levels, and accordingly is one Arrangement of the direction sensing means in these directions is sufficient, or (b) if the main direction of emission is limited to a limited range of angles is limited; For example, the speech direction becomes teleconferences normally limited to a forward angle range.

The Microphones can as a means for direction detection and at the same time as a sound recording device for the Recording the sound signals of the sound source can be used. Based the position of the sound source and possibly also on the basis of the determined The main emission direction can determine a weighting of the microphones be the contribution of each microphone to the recorded audio signal regulates.

The 2A and 2 B show one opposite to the method 1 simplified acoustic method for determining the main emission direction of the sound source.

Instead of the relatively complicated procedure 1 It is also possible to use a much simpler method for determining the main emission direction, which is based on the same principle as in 1 also determines the sound levels in different directions with the appropriate corrections. However, the main emission direction is determined by comparing the detected level ratios in the different directions with a predetermined reference. Is the directional characteristic of the sound source in the form of a measurement, an approximation obtained from measurements, a math matic function, a model or simulation or similar before, it can be used as a reference for determining the main radiation direction. Depending on the complexity of the approximation of the directional characteristic of the sound source selected as the reference, only a few microphones are then necessary to detect the main emission direction. The accuracy and thus complexity of the reference depends on how exactly the main emission direction is to be determined; If a rough determination of the main radiation direction is sufficient, a very simplified reference can be selected. The number and position of the microphones for detecting the sound levels in different directions must be chosen so that the directions scanned therewith are sufficient to unambiguously determine, together with the reference, the location of the directional characteristic of the sound source with respect to the microphones.

For speech signals, for example, assuming a greatly simplified reference for the directional characteristic, as exemplified in FIG 2A shown schematically, the main radiation direction can be determined sufficiently accurately with at least 3, and preferably 4 microphones, which are positioned so that they each enclose an angle of 60 ° -120 °. 2 B shows an example in which the 4 microphones M ₁ to M ₄ each enclose an angle of 90 °.

If the possible main emission directions are restricted to a certain angular range, the in 2A illustrated reference also be further simplified. For example, a backward main radiation direction can be excluded at conferences, if no participants sit behind each other. In this case, the reference of 2A be simplified so that the tip pointing backwards is not taken into account, that is, it is assumed only by an approximately kidney-shaped polar pattern as a reference. In this case, 2 microphones are sufficient, which enclose an angle of 60 ° -120 ° in order to capture the main emission direction with sufficient accuracy. For example, in 2 B the two microphones M ₃ and M ₄ positioned behind the speaker S are saved.

The approximation the directional characteristic of speech with one of the two above Reference pattern proves to be many applications as sufficient, especially for conference applications, in which a relatively rough determination of the main emission direction for one natural Reconstruction is sufficient. For a more precise determination the main emission direction may be in a video conferencing application In addition, the optical means or means used with pattern recognition become. It is also possible the determination of the main emission direction with upstream frequency filters restrict the information-carrying frequency ranges.

As in 1 For example, the microphones intended for direction detection may also be used simultaneously as sound recording means for recording the sound signals of the sound source.

In 3 the determination of the main emission direction of a sound source is shown by means of a reference sound level. The main radiation direction of a sound source T can be determined in a known direction with the aid of a directional characteristic of the sound source, which is the reference, and a current reference sound level of the sound source. Compared to the in 2 explained method can be determined with this method, the main emission even with more complex Referenzrichtcharakteristiken with significantly fewer microphones M. Based on the reference sound level in the known direction, the relative attenuation in the directions predetermined by the microphones M can be determined. Of course, the necessary corrections with regard to the distance of the microphones M to the sound source T and the directional characteristic of the microphones must also be taken into account in this method. The correction may also include knowledge of the geometry of the environment and the related conditions for sound propagation, as well as reflection properties. A comparison of the thus determined relative attenuation with the present as a reference directional characteristic of the sound source T results in the main emission direction.

The reference sound level can be detected, for example, with a clip-on microphone M ₁ , which always follows the changes in direction of the sound source T, so that the direction of the sound signals thus detected is always the same and thus known. It is advantageous if the direction of the reference sound level coincides with the main emission direction. The microphone M ₁ , which is used for the determination of the reference sound level, can be used simultaneously as an acoustic means for recording the sound signals.

For example, if as reference for the directional characteristic of a speech signal in 2A For example, the main emission direction of the sound source can be determined relatively precisely with only 2 direction-detecting microphones M, which enclose an angular range of approximately 60 ° -120 °, and the microphone M ₁ for the determination of the reference sound level.

Also In this method, the determination of the main emission direction with appropriate frequency filters on the information-bearing Frequency ranges restricted become.

In 4 a method for direction detection is shown at several sound sources in the recording room. The individual main emission directions of a plurality of sound sources T ₁ to T ₃ in the recording room are determined by means of a single directional acoustic means associated with all available sound sources.

Are, as in 4 If there are several sound sources T present in the recording space, the determination of the main emission direction of each individual sound source can be carried out using the same methods as described above for a single sound source. For this purpose, however, the sound signals of the individual sound sources T _x must be present separately for their direction detection. This is automatically the case when only one sound source emits sound at a given time. However, if two or more sound sources emit sound at the same time, the sound signals of the individual sound sources, which are all received simultaneously from the direction sensing means by the microphones M ₁ to M ₄ , have to be separated from each other by a suitable method for their direction detection. The separation can be done, for example, with a conventional source separation algorithm. It is particularly easy to assign the sound signals to the corresponding sound sources if the separate sound signals of the sound sources are known as reference signals. These reference signals are obtained, for example, if for the recording of the sound signals for each sound source separately an acoustic means, for. As a microphone M _T1 , M _T2 and M _T3 , is used, as in 4 shown. All sound signals that do not belong to the assigned sound source whose main emission direction is to be determined are suppressed for the direction determination. The separation of the sound signals by means of the reference signals can be improved by also taking into account the different transfer functions that apply to the microphones of the direction detecting means (M, to M ₄ ) and to the means (M _T1 , M _T2 and M _T3 ).

In the in 4 example shown, the separate detection of the main emission of the individual sound sources with a direction sensing means according to the in 2 illustrated method. As stated therein, the direction sensing means may consist of 4 microphones, including an angular range of about 60 ° -120 °; but it is also possible to use only the 2 microphones placed in front of the participants.

In 5 a method is shown in which a separate directional acoustic means is used for each sound source. For detecting the main emission directions of a plurality of sound sources T ₁ to T ₃ in the recording room, each sound source can be assigned its own direction-detecting means M ₁ to M ₃ . Since each sound source has its own directional detection acoustic means, this type of process eliminates the need to separate the sound signals from the associated sound sources. At the in 5 the example shown, the main emission of each sound source with the im 2 determined method determined. Since in many conference applications, especially in videoconferencing, a backward speech direction can usually be ruled out, 2 microphones are sufficient to determine the main radiation direction of a sound source with sufficient accuracy.

The sound signals of the sound sources are recorded in 5 optionally with an additional microphone M ₁ 'to M ₃ ' per sound source, which is assigned to each sound source T ₁ to T ₃ , or the direction detecting microphones M ₁ to M ₃ are also used simultaneously for recording the sound signals.

In 6 there is shown a sound source reproduction method comprising a first reproduction unit and at least a second spaced reproduction unit.

The in the recording room recorded sound signals TS of a sound source can in Play room with a first playback unit associated with the sound source WE1 are played. The position of the first playback unit WE1 can be chosen like this be that they are the virtual position of the sound source in the playback room equivalent. In a videoconference, this virtual position can for example, at the point in the room where the visual representation the sound source is located.

For conveying the direction information of the sound reproduction, at least one second reproduction unit WE2 spaced from the first reproduction unit is used. Preferably, two second display units are used, one of which can be positioned to one side and the other to the other side of the first display unit WE1. With such a structure, changes in the main emission direction of the sound source can be simulated in an angular range of 180 ° around the first reproduction unit, that is, the virtual sound source positioned at that position. The information about the emission direction can be imparted by the fact that the reproduction with the second reproduction units is delayed with respect to the first reproduction unit. The time delay τ should be selected such that the actual time delay Δt = t _WE2 -t _WE1 between the audio signals is at least in subregions of the playback _room between 2 ms and 100 ms, thus at the receivers, ie, for example, at the receiving participants of the video conference , the are in these subregions, the actual time delay is between 2 ms and 100 ms.

The The main emission direction HR detected in the recording room controls the reproduction levels at the second playback units via an attenuator a. For example to simulate a main emission direction of the sound source that is on The right side of the room is addressed, the sound signals to the second playback unit, which is on the left, fully muted and only over the right second reproduction unit delays with respect to the first reproduction unit played.

The Of course, the above-described method can also be applied to several Sound sources are used in the recording room. For this purpose must be accordingly more first and second playback units are used.

The 7A and 7B show various methods for the realization of the first and second display units.

The first as well as the second display units WE1 and WE2 can, as in 7A each realized with a real speaker or a group of loudspeakers at the corresponding position in the room. However, they can also be realized in each case with a virtual source, which is placed at the corresponding position, for example by wave field synthesis, as in 7B shown. Of course, a mixed realization with real and virtual sources is possible.

In the 8A and 8B For example, there is shown a sound source reproduction method having a first display unit and a plurality of second, spaced-apart display units.

This in 6 explained basic method can be extended with the additions explained below to reproduce the direction information of the sound source as closely as possible.

On the one hand, it is possible, instead of in each case one second reproduction unit on each side of the first re-unit WE1, to use a plurality of spaced-apart second reproduction units WE2, as in FIG 8A shown. The delays τ to the individual reproduction units WE2 can be selected individually for each reproduction unit. For example, it is particularly advantageous with increasing distance of the playback units WE2 from the playback unit WE1 to select the corresponding delays shorter. This must be how to 6 It should be noted that the actual time delay between the audio signals, at least in subregions of the playback room, is between 2 ms and 100 ms, preferably between 5 ms and 80 ms, and in particular between 20 ms and 40 ms.

As in 8A illustrated, according to the directional characteristic of the sound source to be simulated, the sound signal TS before playback by the second (s) playback unit (s) WE2 additionally be processed with a filter F, for example a high-pass, low-pass or band-pass filter.

For as possible Perceptive reproduction of the emission direction can also the Playback level of the first and second playback units depending on be adapted to be simulated directional characteristic. The playback levels be this by means of an attenuator a set so that the resulting from the directional characteristic perceptible loudness differences at different listener positions Approached accordingly become. The thus determined attenuations for the individual playback units can for different Main radiation directions HR are defined and stored. at a sound source with time-varying directivity controls then the detected main emission direction the reproduction levels of individual playback units.

In 8B schematically shown as an example, the attenuation functions for a first and two second playback units on each side of the first playback unit (WE1, WE2 _L1 , WE2 _L2 , WE2 _R1 , WE2 _R2 ) in dependence of the main emission direction HR, as they are for the control of the directional playback can be stored. For the sake of simplicity, instead of the logarithmic level values, the sound pressure of the corresponding reproduction unit is shown in relation to the sound pressure of the sound signal p _TS . Depending on the detected and transmitted Hauptabstrahlrichtung HR the attenuators a of the respective playback units are set according to the stored preset. In the example chosen, care must be taken that for each possible main emission direction the level value of the first reproduction unit is either greater than or equal to the corresponding level values of the second reproduction units, or at most 10 dB, or better 3 to 6 dB, smaller than the corresponding level values of the second Is playback units.

The Of course, the above-described method can also be applied to several Sound sources are used in the recording room. For this purpose must be accordingly more first and second playback units are used.

In 9 There is shown a multi-sound source reproduction method with overlapping first and second display units.

If there are several sound sources in the recording room, the sound signals of the sound sources, such as the 6 and 8th explained, are played with first and second playback units in the playback room. However, the number of necessary playback units can be significantly reduced, if not for each sound source own first and second playback units provides. Instead, the playback units can simultaneously serve both as first and second playback units for different sound sources. It is particularly advantageous to associate with each sound source a first playback unit which is located at the virtual position of the respective sound source in the playback room. As the second reproduction units for one sound source, the first reproduction units of the adjacent sound sources can then be used. In addition, it is also possible to use display units which serve exclusively as second playback units for all or at least part of the sound sources.

In 9 For example, there is shown an example of four sound sources in which each sound source is assigned a first playback unit and on each side of the first playback unit, with two exceptions, two second playback units. The sound signals TS1, TS2, TS3 and TS4 of the four sound sources are reproduced with their associated first reproduction units WE1 placed at the corresponding virtual positions of the sound sources in the reproduction room. The first reproduction units WE1 also serve as second reproduction units WE2 for the adjacent sound sources at the same time. The time delays τ _{1 of} these second display units are preferably chosen so that the actual time delays between the audio signals are at least in subregions of the playback room between 5 ms and 20 ms. In addition, two second reproduction units WE2 'are provided in this example, which serve exclusively as second reproduction units for all four sound sources. The time delays τ _{2 of} these second display units are set such that the actual time delays between the audio signals at the receivers ie at the receiving subscribers of a video conference, in the playback room are between 20 ms and 40 ms.

As in 8th That is, the main emission directions HR of the sound sources detected in the recording space control the reproduction levels of the first and second reproduction units via the respective attenuators a. Of course, it is also possible to process the audio signals additionally with a filter F, wherein the filter can be selected individually for each audio signal or for each playback unit WE2 or WE2 '. Since the number of accumulated sound signals reproduced through a reproducing unit may vary, it is advantageous to normalize the reproduced level corresponding to the current number with a normalizing term NOM.

The 10A and 10B demonstrate a simplified reproduction method for direction detection 5 , Each sound source is assigned its own directional acoustic means.

How to 5 1, a direction detecting means may be associated with each sound source for detecting the main radiation directions of a plurality of sound sources in the recording room. In this case, the reproduction of the emission directions-by means of first and second reproduction units-can take place directly with the tone signals of the corresponding sound source detected in different directions. In the following exemplary embodiment, this reproduction method is explained by way of example with reference to a sound source. For multiple sound sources, the method must be extended according to the same principle, wherein the technique of overlapping display units explained in Example 9 can be used to reduce the necessary number of first and second display units.

In 10A the sound source is shown with its associated means for detecting the main emission direction and the optional microphone for recording the sound signal TS in the recording room. To detect the emission direction, four microphones are used in this example, which record the audio signals TR ₉₀ , TR ₄₅ , TL ₉₀ and TL ₄₅ . For the recording of the sound signal TS of the sound source either a separate microphone can be provided, or the sound signal is formed from the recorded sound signals of the direction detecting means in the reproduction, as in 10B shown.

In 10B the reproduction method is shown by means of first and second reproduction units. To convey the direction information, the sound signals TR ₉₀ , TR ₄₅ , TL ₉₀ and TL ₄₅ recorded with the direction-sensing means are reproduced via the corresponding second reproduction units WE2 delayed to the sound signal TS. The time delays τ can be selected as explained in the preceding examples. Since the directional level differences are already included in the recorded sound signals of the direction detecting means, the level control of the second display units is not necessary by the main emission direction; the attenuators a are therefore only optional. The sound signals can additionally be processed by the second reproduction units WE2 with a filter F in accordance with the directional characteristic to be simulated before the reproduction.

The reproduction of the sound signal TS of the sound source via the first playback unit. In this case, the sound signal TS can either be the sound signal recorded with its own microphone, or it can be formed from the sound signals TR ₉₀ , TR ₄₅ , TL ₉₀ and TL ₄₅ , for example by using the largest of these sound signals or the sum of the four sound signals. In 10B is shown as an example the summation.

The sound quality of the explained Reproduction method may indeed be affected by comb filter effects; Nevertheless, due to its simplicity, the method can be used in some Applications can be quite beneficial.

Claims (19)

Method for recording sound signals from a sound source with a time-variable directional characteristic arranged in a recording room with sound recording means and for reproducing the sound signals with sound reproduction means in a reproduction room, characterized in that the main emission direction of the sound signals emitted by the sound source is detected as a function of time and the Playback in dependence of the detected main emission takes place. Method according to claim 1, characterized in that that the sound signals of the sound source are recorded by a sound recording medium be, and the main emission of the emitted sound signals is detected by means for direction detection. Method according to claim 2, characterized in that in that the means for direction detection are of the acoustic type, in particular Microphones and / or one or more microphone arrays. Method according to claim 2 or 3, characterized that the means for direction detection are optical, in particular a video capture with pattern recognition. Method according to one of claims 1 to 4, characterized that the reproduction of the sound signals associated with one of the sound source first playback unit and at least one spaced from the first playback unit arranged second playback unit he follows. Method according to claim 5, characterized in that that the position of the first playback unit in the playback room corresponds to a virtual position of the sound source in the playback room. Method according to claim 5 or 6, characterized that playback with the second or the second playback units across from the first playback unit with time delays τ takes place. Method according to one of claims 5 to 7, characterized the time delay τ is chosen such that that time delay between the audio signals at least in subregions of the playback room between 2 ms and 100 ms, preferably between 5 ms and 80 ms and especially between 10 ms and 40 ms. Method according to one of claims 5 to 8, characterized that the reproduction by the first and / or by the second (s) Reproducing unit (s) with a reduced level, in particular reduced by 1 to 6 dB and preferably by 2 to 4 dB Level, and / or in particular in dependence on the main emission, takes place. Method according to one of claims 1 to 9, characterized that the playback units speakers or a group of speakers, a speaker array or a combination thereof or a virtual one Source, in particular a generated by wave field synthesis virtual Source, are. Method according to one of claims 1 to 10, characterized that sound signals from a plurality of sound sources arranged in the recording room recorded and played in the playback room. Method according to claim 11, characterized in that that each sound source is assigned sound recording means. Method according to claim 12, characterized in that that sound signals from a sound source that is not assigned to the sound source Receiving means are received by an echo cancellation or crosstalk suppression (cross talk cancellation) become. Method according to one of claims 1 to 13, characterized that the playback is done by wave field synthesis. System for recording sound signals from one or more sound sources with time-varying directional characteristics with sound recording means in one in a recording room and for reproducing the sound signals with sound reproduction means in a reproduction room, characterized in that the system comprises means for detecting the main emission directions of the sound signals emitted by the sound source (s) as a function of time and means for reproducing the transmitted sound signals as a function of the detected directions , System according to claim 15, characterized in that the system has at least two sound recording units assigned to a sound source for recording the sound signals emitted by this sound source and having the main emission. System according to claim 15, characterized in that the system comprises at least one sound recording unit associated with a sound source for recording the sound signals emitted by this sound source and having optical means for detecting their Hauptabstrahlrichtung. System according to one of claims 15 to 17, characterized that the number of sound recording units and / or sound reproduction units the number of sound sources plus 2 corresponds. System according to one of Claims 15 to 18, characterized that the sound reproduction units are speakers or a group of Speakers, a speaker array or a combination thereof or a virtual source, in particular one by wave field synthesis generated virtual source, are.